2020 Chatbot Trends: Five Takes on the Future of The Industry
A new year, a new decade. We're in the future now.
We should be tapping the badge on our chests and talking to 'computer' very soon. Right?
Instead of finger-in-the-air predictions from some mid-tier marketer (me), I wanted to research and find out from the people buying, using and building these things what happened with chatbots in 2019 and what they are going to be doing with the technology in 2020.
So, I cornered five people and asked them...
"What was the biggest impact of chatbots in 2019, and what are you planning to do with the technology in 2020?"
Response one
ubisend customer, an HR service manager
HR departments in medium to the enterprise is our largest customer segment. It's no surprise really, it's a good starting point for these companies to test conversational software -- it solves a real problem.
"We started working with ubisend on our HR help desk chatbot in July last year. We didn't have our service processes and knowledge ready for automation so the initial launch took a bit longer than planned. Our HR chatbot is now live and responding to staff. Launching this chatbot forced us to centralise our knowledge. Now, the biggest impact we see is a reduction in inbox traffic from employees to our HR reps.
We're switching to SuccessFactors in Q3. Our goal in 2020 for the chatbot is to build it directly into the employee-facing side of the new HR system."
Response two
ubisend chatbot developer
Time to get a little geeky now.
I managed to encourage one of our developers to speak to another human, as I wanted a bit of insight from someone who builds these things.
Interestingly, he was more interested in the tech hype than the code stuff.
"At the start of 2019, it was all about using advanced natural language processing (NLP) and machine learning (ML) to enable the user to free-talk with a computer. Note from the editor: read up on the key lingo with our chatbot glossary.
While this sounds (and is) fun, it didn't exactly work in business-focussed bots. When someone wants help from a machine they don't want to talk about the weather or engage in small talk, they want an answer.
Towards the end of 2019, the eureka moments came when chatbots focussed on efficient and more structured, RPA-based, experiences. It became less about 'hey look you can talk to a computer, how fun!' and more 'I've got this problem, computer, fix it'.
While small-talk is still fun and an essential part of the experience, it should never be the reason for a business-focussed chatbot to exist.
If 2019 was the year of voice (it kind of was in consumer solutions but not really in business), then 2020 will be the year formalising what voice chatbots can, and should, do. At the moment we're still in the testing phase of voice solutions. No one has figured out what the best use case for them is, either in business or consumer-facing solutions.
The big impact this year will be when voice becomes the natural way of (efficiently) getting information from a computer."
Response three
ubisend customer, head of sales and marketing
As could probably have been expected, my conversation with a customer who purchased a custom sales chatbot centred around metrics and results.
Their goal was for the chatbot to increase the conversion rates on one of their e-commerce websites, so their 2019 impact and 2020 predictions were rather numbers based!
"For us, the big lesson we learned in 2019 was that tech can be a relevant tool in the sales processes.
We added an onsite popup chat widget chatbot which we wanted to capture interest and move somebody through the buying journey. Customer feedback has been excellent, with over 75% of users rating their experience five stars. The chatbot became a valuable lead-generating source.
2020 will be all about testing. Now the solution is built, it's cost-effective for us to use it to test other channels, promotions, and products. Our chatbot can be connected to all our social channels and email at once and simultaneously test 24 hours per day.
Rather than my team having to build campaigns and monitor performance, the chatbot will subtly change its language, sales flow and product recommendations at scale.
As we go through 2020, our chatbot will produce vast amounts of both qualitative and quantitative data, enabling us to be agile in sales and marketing processes and content."
Response four
ubisend chatbot implementation specialist
A ubisend implementation specialist is a highly trained, geeky and techie project manager. They know conversational technology inside and out, and have learned what works and what doesn't. They are the interface between our customer and our geeks.
This is our chatbot development team structure. We're talking about the project manager, in the top right!
With their finger on so many pulses, they offer valuable insight on 2019 impact and 2020 predictions.
"I think 2019 was all about expectations. All the hype and over-inflated sales and marketing content produced by the chatbot development industry (not ubisend, of course!) meant customers had unattainable expectations of the technology.
They expected their solution to go live, and instantly be able to answer all questions 100% accurately and provide a perfect service.
Unfortunately, the way artificial intelligence and machine learning work means that both us, as the development company, and the customer have to put a certain amount of effort into training the chatbot. With current technology, it's impossible to launch a new solution with human-level accuracy; it takes a little time to make it smart.
2020 is going to be exciting. It's going to be the year everyone realises chatbots aren't gimmicks, and they can do what they're designed to do.
More evidence from real-world deployments will continue to emerge, and conversational software will very much become the norm across both consumer and business communication.
I'm particularly excited about e-commerce and dealing with consumer pre-order questions. Consumers expect businesses to be available all the time.
Note from the editor: 25% UK consumers expect businesses to answer within ten minutes (source).
If they don't get an answer immediately, they'll take their money somewhere else. The website or social media-based chatbot effectively, and accurately, answering pre-order questions 24 hours a day in any language and on any channel is going to be huge."
Response five
ubisend chatbot salesperson
After speaking to one of our SDRs, I was amazed at how much they talked about humans and not technology.
They spoke of saving time, driving efficiency, and enabling people to use the technology effectively. Not once did they speak of the nuts and bolts of what it is we fundamentally build.
"2019 was all about driving efficiencies, understanding what can and what shouldn't be automated and developing custom chatbot software to solve problems.
It didn't seem to matter how smart the solution was, what data it connected to, or the programming language used to build it.
What mattered to our customers was the positive impact it had on people.
2020 is going to be all about reducing workload. The technology will stop humans doing the simple and 'dumb' things a computer can do (faster and without stopping).
More case studies of successful implementations will appear (from our customers and others), and we're going to see mass-adoption of efficient, systematic and task-orientated chatbots by companies of all sizes and industries.
In 2020, we're going to see chatbots enabling humans to spend their time on critical tasks while the technology deals with all the mundane, routine tasks."
How to prepare for 2020 and beyond
As companies strive to adapt, evolve and implement new technologies, the chatbot implementation space will only get more competitive.
The new year (and new budgets?) provides an opportunity for leaders in HR, customer service, and sales to leap-frog the competition and take the lead within their industry with the most efficient, reliable and cost-effective intelligent automation.
Talk to one of our team today and see if your human-problems can be solved with machine efficiency.
